Skip to content

DOME Recommendations for supervised ML validation in biology

DOME image

Motivation

Modern biology frequently relies on machine learning (ML) to provide predictions and improve decision processes. There have been recent calls for more scrutiny on ML performance and possible limitations.

What is DOME-ML?

DOME-ML (or simply DOME) is an acronym standing for Data, Optimization, Model and Evaluation in ML. DOME is a set of community-wide guidelines, recommendations and checklists spanning these four areas aiming to help establish standards of supervised ML validation in biology. The recommendations are formulated as questions to anyone wishing to pursue implementation of a ML algorithm. Answers to these questions can be easily included in the supplementary material of published papers.

What is the scope of the recommendations?

The recommendations cover four separate aspects covering the major areas of ML:

  • Data
  • Optimization
  • Model
  • Evaluation

About this training course

Learning and using the DOME recommendations will increase awareness among researchers on best practices to share ML approaches so they include minimum metadata for other researchers to get a quick and clear picture of the their ML approach is about and how it compares to others. This course aims to be a useful and straightforward guide for those who wish to incorporate DOME in their work.

Authors

Fotis E. Psomopoulos

Fotis E. Psomopoulos

:simple-linkedin: 🌍
Senior Researcher(INAB|CERTH)

Styliani-Christina Fragkouli

Styliani-Christina Fragkouli

:simple-linkedin: 🌍
PhD Candidate(INAB|CERTH)

Lesson overview

Description
The DOME recommendations are thoroughly presented. Every section is well described in order for all learners to be able to implement them.

Prerequisites
To be able to follow this course, learners should have knowledge in:
 1. ML basic concepts
 2. Second requirement

Learning Outcomes:
By the end of the course, learners will be able to:
 1. Recognize the DOME recommendations
 2. Implement the DOME recommendations

Target Audience: Anyone who uses ML in their research.

Level: Beginner

License: Creative Commons Attribution 4.0 International License

Funding: This project has received funding from [name of funders].

Contributors

Prometheus

Prometheus

:simple-linkedin:
Lorem ipsum dolor sit amet.

Citing this lesson

Please cite as:

  1. Enter your citation here.
  2. Geert van Geest, Elin Kronander, Jose Alejandro Romero Herrera, Nadja Ε½lender, & Alexia Cardona. (2023). The ELIXIR Training Lesson Template - Developing Training Together (v1.0.0-alpha). Zenodo. https://doi.org/10.5281/zenodo.7913092.

Setup

Software setup

To run this course you just need to have an internet browser.